Simple to-do app using AWS API Gateway and AWS Lambda. This is the sample code to the detailed walkthrough on error handling in a serverless architecture.

Note: This project is mainly for demonstration purposes and includes only the backend of a typical to-do app – limited to add and get single tasks. Also, the implementation uses a mock store and doesn't store tasks persistently.

Feel free to play around with the sample code to extend the functionality or add an implementation for persistent storage, such as DynamoDB, PostreSQL or others.

See below for how test the API using curl or use a HTTP client of your choice.

You'll need the following tools to build, run and deploy the app. In order to deploy the app you must have access to an AWS account.

• AWS Command Line Interface
• AWS SAM Local
• Docker (Mac, Windows, see distro package manager on Linx)

The app has two commands, one to add and another to get a task. Each is a separate Lambda function and needs to be built separately.

./cmd/add-task/scripts/build.sh

./cmd/get-task/scripts/build.sh

Each command comes with example events and uses example-event.json by default. To specify the file use -e <PATH_TO_EVENT> when calling the invoke script.

./cmd/add-task/scripts/invoke.sh

Note: Before deploying make sure to build both commands, see above.

In order to deploy the full stack, package the CloudFormation template using the following command. Make sure to set S3_BUCKET correctly by specifying the name of the bucket used to store all build artifacts – create a bucket if necessary.

aws cloudformation package \
    --template-file cloudformation.yaml \
    --s3-bucket <S3_BUCKET> \
    --output-template-file packaged-template.yaml

The packaged template can then be used to bring up the CloudFormation stack.

aws cloudformation deploy \
  --template-file packaged-template.yaml \
  --stack-name <STACK_NAME> --capabilities CAPABILITY_NAMED_IAM \
  --parameter-overrides StageName=<STAGE_NAME>

Replace the API_GATEWAY, AWS_REGION and STAGE_NAME in the cURL commands with your own values. You can find the API Gateway identifier in the console within the Dashboard of your API and where it shows the Invoke URL.

curl -X "POST" "https://<API_GATEWAY>.execute-api.<AWS_REGION>.amazonaws.com/<STAGE_NAME>/tasks" \
    -H 'Content-Type: application/json' \
    -d $'{
  "note": "Take the umbrella ☔"
}'

{
  "id": "2b837aa9-9999-40b8-bc0c-42536f7272f7",
  "done": false,
  "note": "Take the umbrella ☔",
  "created_at": "2018-04-20T13:37:34.249306491Z",
  "updated_at": "2018-04-20T13:37:34.249306491Z"
}

curl -X "POST" "https://<API_GATEWAY>.execute-api.<AWS_REGION>.amazonaws.com/<STAGE_NAME>/tasks" \
    -H 'Content-Type: application/json'

{
  "code": "BAD_REQUEST_BODY",
  "message": "Invalid request body"
}

curl -X "POST" "https://<API_GATEWAY>.execute-api.<AWS_REGION>.amazonaws.com/<STAGE_NAME>/tasks" \
    -H 'Content-Type: application/json' \
    -d $'{
  "note": ""
}'

{
  "code" : "INVALID_INPUT",
  "message" : "Invalid input"
}

curl -X "POST" "https://<API_GATEWAY>.execute-api.<AWS_REGION>.amazonaws.com/<STAGE_NAME>/wrong-path" \
    -H 'Content-Type: application/json' \
    -d $'{
  "note": "Take the umbrella ☔"
}'

{
  "code": "MISSING_AUTHENTICATION_TOKEN",
  "message": "Missing Authentication Token"
}